Prospects for gene therapy of neutrophil defects.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to H L Malech.
Explore the source record for details and available documents.
Infection is a major cause of morbidity and mortality in patients after thermal injury. This predisposition to infections is related, in part, to abnormal polymorphonuclear leukocyte (PMN) function and a diminished respiratory burst. To evaluate the biochemical basis for the defective respiratory burst after major burns, the status of the oxidase enzyme system and its components was investigated. PMNs were isolated from 24 patients with 12% to 62% burns. Oxidase activity of intact PMNs, measured as superoxide anion (O2-) generation or oxygen consumption, was decreased in burn compared with healthy controls. Subcellular fractions from patient PMNs generated less O2- in the sodium dodecyl sulfate cell-free system, and this was related to a diminished contribution by cytosol but not by plasma membrane. Subsequently, cytosol was separated with CM-Sepharose, yielding two fractions; one contained the p47-phox and p67-phox (47/67 mix) and the other contained the remaining cytosolic components (run through [RT]). Although the contribution to oxidase activity made by RT from patient cytosol was similar to that of control, the activity of p47/67 mix from PMNs of burn patients was deficient. Quantitative assays using an immunoautoradiographic technique showed a consistent, but significant decrease in both p47-phox and p67-phox. The addition of purified or human recombinant p47-phox but not p67-phox corrected the diminished oxidase activity of cytosol from burn patients. Thus, decreased respiratory burst activity found in PMNs from individuals with thermal injury was associated with a specific, quantitative deficiency of p47-phox.
Cellular iron metabolism is altered during chronic inflammatory states, leading to reticuloendothelial iron sequestration and an associated anemia. To study the effects of nitric oxide (NO) on the expression of three genes involved in erythroid cell iron metabolism (gamma-globin, H-ferritin, and transferrin receptor [TfR]), we developed a series of human K562 erythroleukemic cell clones retrovirally transduced with inducible nitric oxide synthase (NOS-2) and producing different steady-state levels of NO. gamma-Globin and H-ferritin protein expression was reduced in NO-producing cells in relation to the amount of NO produced. Conversely, cell surface TfR expression increased in NO-producing clones. Both the inhibitory effects of NO on gamma-globin and H-ferritin expression and the stimulatory effect on TfR were reversed by the NOS inhibitor NG-methyl-L-arginine (NGMMA). gamma-Globin and H-ferritin mRNA levels were unaffected by NO production. In the case of TfR, NO appeared to stabilize mRNA in that the half life of TfR mRNA decreased from approximately 15 hours to less than 3 hours when NO production by NOS-transduced clones was inhibited. Thus, NO can regulate expression of these genes at the posttranscriptional level, an effect that is likely mediated by the known effect of NO on the RNA binding activity of iron regulatory protein-1 (Pantopoulos and Hentze, Proc Natl Acad Sci USA 92:1267, 1995). Furthermore, our findings suggest a mechanism for the observed relationship between NO production and the pathophysiology of the anemia of chronic disease.
Recent studies have indicated that nitric oxide may affect iron metabolism through disruption of the iron-sulfur complex of iron regulatory protein-1, a translational regulator. In the present study, we report that heterologous expression of murine macrophage nitric oxide synthase (NOS-2) in the human erythroleukemic K562 cell line results in constitutive production of nitric oxide associated with inhibition of hemoglobin expression. K562 cells were transfected with an episomally-maintained, hygromycin-selectable expression vector bearing the coding region of NOS-2. Constitutive NOS expression was detected by Western blotting of cell lysates and by the accumulation of nitrite in the culture media. Although NOS-transfected cells grew more slowly than control cells, they were able to maintain constitutive expression of NOS and production of nitric oxide for more than 1 month following transfection. The hemoglobin content of NOS-transfected K562 cells was less than one-fifth that of control cells, but increased markedly if NOS inhibitor was included in the culture media. The nitric oxide-mediated inhibition of hemoglobin expression was reversed by supplementing the culture media with 20 mumol/L hemin or 0.5 mmol/L 5-amino-levulinate, indicating that nitric oxide did not directly inhibit hemoglobin synthesis, but likely acted on a step in heme synthesis. mRNA levels for globin and erythroid aminolevulinic acid synthase (eALAS) were the same in both NOS-transfected and control cells. Our observations indicate that hemoglobin expression is inhibited by nitric oxide in NOS-transfected K562 cells by posttranscriptional repression of eALAS, the first enzyme of the heme biosynthetic pathway. The most likely mechanism is a nitric oxide-mediated translational repression of eALAS, as was recently demonstrated for ferritin synthesis. These observations further illustrate the potential for endogenously produced nitric oxide to regulate cellular posttranscriptional events. In particular, our observations may be relevant to the role of nitric oxide in anemia and lowered blood hemoglobin concentrations that are associated with chronic infections, such as tuberculosis or parasitic disease.
Peripheral blood (PB) CD34+ cells mobilized by granulocyte colony-stimulating factor (G-CSF) administration are potentially useful for transplantation and as a target of gene transfer for therapy of hematopoietic disorders. Efficient harvest and planning for clinical use of PB CD34+ cells ideally requires foreknowledge of the expected mobilization kinetics and yield. We developed a sensitive flow cytometric assay for accurately enumerating CD34+ cells throughout the range seen at baseline to peak mobilization. We used this assay to assess the kinetics of G-CSF-mediated mobilization of CD34+ cells to PB in normal volunteers and in patients with chronic granulomatous disease (CGD) or adenosine deaminase (ADA)-deficient severe combined immunodeficiency disease (SCID). Two dose levels of G-CSF were examined (5 and 10 micrograms/kg/d for 7 days). Both doses were well tolerated. For normal subjects and patients an increase in PB CD34+ cells was first detected only preceding the third dose of G-CSF (day 3), peaked transiently on day 5 or 6, and then decreased thereafter despite additional doses of G-CSF. With 32 normal volunteers mean peak CD34+ cell counts were 57 and 76 cells/mm2 of blood (5 and 10 micrograms doses, respectively), whereas for 18 CGD patients the mean peaks were 31 and 40 cells/mm2 of blood. For 2 ADA-deficient SCID patients studied at a G-CSF dose of 5 micrograms/kg/d, the average peak was 16 cells/mm2 of blood. For both of these patient groups mobilization of CD34+ cells to PB was impaired compared with similarly treated normal subjects (P < .05). By contrast to the kinetics of the CD34+ cell mobilization, the absolute neutrophil count (ANC) increased markedly by 6 hours after the first dose of G-CSF and then increased steadily through day 8. At days 5 and 6 (peak mobilization of CD34+ cells) the mean ANC of CGD and ADA patients was only slightly lower ( < or = 15%) than that seen with normal subjects, whereas the difference in CD34+ cell mobilization was > 48%. Thus, ANC is not a reliable surrogate to predict peak PB CD34+ cell counts and direct enumeration of PB CD34+ counts should be undertaken in decisions regarding timing and duration of apheresis to harvest a specific number of these cells. Finally, unexpected, but significant differences in the PB CD34+ cell mobilization between normal subjects and patients with inherited disorders can occur and underscores the importance of establishing the expected mobilization of PB CD34+ cells in the planning of treatment approaches using these cells.
Immunomodulatory cytokines have been used with success as adjunctive therapy in genetic disorders such as chronic granulomatous disease and infectious diseases such as leishmaniasis and leprosy. As the first step toward developing novel methods to deliver immunomodulatory cytokines, we used retrovirus-mediated somatic gene transfer techniques to produce IFN-gamma from human peripheral blood CD34+ hemopoietic progenitor (PBHP) cells. After transduction, the PBHP cells were made to differentiate toward myelo-monocytic lineages. Only the PBHP-derived myelo-monocytic cells that were transduced with the IFN-gamma cDNA produced IFN-gamma(4 +/- 1.3 ng of IFN-gamma/10(6) PBHP cells.) Despite a reduction in the proliferation of IFN-gamma-transduced PBHP cells as well as a decrease in erythroid colony formation, there was an enhancement of monocyte differentiation and activation. Monocytes differentiated from the IFN-gamma-transduced PBHP cells demonstrated 1) up-regulation of MHC class I and II Ag expression, 2) increased Fc(gamma)RI expression, and 3) enhanced superoxide production in response to both opsonized zymosan (25-fold) and phorbol ester (3-fold). Furthermore, a functional response to a monocyte-specific chemokine, monocyte chemotactic protein-1 (mobilization of intracellular Ca2+) was seen only in the IFN-gamma-transduced cells. Thus, PBHP cells transduced with IFN-gamma cDNA produce not only biologically active IFN-gamma, but also enhanced monocyte differentiation, resulting in an activated state that includes unique functions, such as responsiveness to monocyte chemotactic protein-1. These transduced activated monocytes may be specifically suited to cellular therapy requiring homing to sites of inflammation where their anti- microbicidal, cytotoxic and APC functions play an important role in host defense against foreign pathogens.
Gene therapy for inherited disorders of blood cells will require both efficient methods for stable gene transfer and nonablative bone marrow conditioning regimens to allow engraftment of modified hematopoietic progenitor cells (HPCs). We have used a sensitive murine system for detecting HPC engraftment using congenic C57BL/6 mice that differ at the Ly5 locus, which encodes the leukocyte common antigen. The system relies on the ability of monoclonal antibodies with specificity for Ly5.1 and Ly5.2 (revised nomenclature: CD45.1 and CD45.2, respectively) to distinguish donor and recipient peripheral blood leukocytes after transplantation of purified Sca-1+ bone marrow-derived HPCs. No detectable engraftment occurred in nonirradiated recipient mice, even when as many as 2.0 x 10(6) SCa-1+HPCs were transplanted. However, in mice receiving total body irradiation (TBI), engraftment increased as a function of pretransplantation radiation dose, number of transplanted cells, and time after transplantation. Moreover, mice receiving either granulocyte colony-stimulating factor (G-CSF) or G-CSF+ stem cell factor before low-dose TBI (160 cGy) exhibited a marked increase in engraftment compared with mice receiving a vehicle control before low-dose TBI (18.9% and 20.6% v 5.6% at a 1 month, respectively; 29% and 35% v 15.1% at 4 months, respectively). Use of growth factor pretreatment even allowed TBI doses as low as 30, 70, or 120 cGy to achieve significant engraftment of donor progenitors (0.3%, 1.5%, and 6.8% at 1 month, respectively; 1.7%, 5.8%, and 13.9% at 4 months, respectively). All animals remained healthy during the observation periods. Thus, growth factor preconditioning of the recipient followed by low-dose TBI may provide an optimal balance between safety and efficacy in achieving required levels of engraftment for gene therapy of blood disorders.
Explore the source record for details and available documents.
Nitric oxide synthases (NOSs) represent a special case of the cytochrome P450: cytochrome P450 reductase system in which the two redox partners occur as distinct domains on the same polypeptide chain and are linked by a calmodulin binding regulatory sequence. We expressed the carboxy-terminal, flavoprotein domain (residues 527-1144) of murine NOS2 in E. coli. The UV-visible spectrum of this domain resembles those of other flavoproteins, and the protein catalyses the reduction of ferricytochrome c by NADPH [Vmax = 3.1 +/- .1 mol cytochrome reduced/minute/mmol flavoprotein domain, Km (cytochrome c) = 23 +/- 2 microM, Km (NADPH) 0.30 +/- .06 microM]. The high reductase activity of this NOS2 flavoprotein domain, which lacks a calmodulin binding site, precludes a role for this site in mediating electron transfer within the flavoprotein domain of the intact enzyme. This is in contrast to the case of NOS1 and suggests that electron transfer is regulated differently in the two isoforms.
Src homology 3 (SH3) domains mediate specific protein-protein interactions crucial for signal transduction and protein subcellular localization. Upon phagocyte stimulation, two SH3 domain-containing cytosolic components of the NADPH oxidase, p47phox and p67phox, are recruited to the membrane where they interact with flavocytochrome b558 to form an activated microbicidal oxidase. Deletion analysis of p47phox and p67phox in transfected K562 cells demonstrated multiple SH3-mediated interactions between p47phox and the transmembrane flavocytochrome b558 and also between the cytosolic components themselves. The core region of p47phox (residues 151-284), spanning both SH3 domains, was required for flavocytochrome-dependent translocation and oxidase activity in whole cells. Furthermore, translocation of p67phox occurred through interactions of its N-terminal domain (residues 1-246) with p47phox SH3 domains. Both of these interactions were promoted by PMA activation of cells and were influenced by the presence of other domains in both cytosolic factors. Deletion analysis also revealed a third SH3 domain-mediated interaction involving the C-termini of both cytosolic factors, which also promoted p67phox membrane translocation. These data provide evidence for a central role for p47phox in regulation of oxidase assembly through several SH3 domain interactions.
Chronic granulomatous disease (CGD) is an inherited hematologic disorder involving failure of phagocytic cell oxidase to produce superoxide (O2-.), resulting in recurrent infections. The success of retrovirus gene therapy for hematopoietic diseases will be limited both by the efficiency of ex vivo transduction of target cells and by the ability of corrected cells to replace uncorrected cells in vivo. Using MFG-based retrovirus vectors containing oxidase genes, we have previously demonstrated in vitro correction of CGD, but transduction rates were low. In the present study we explore a strategy for providing a selective growth advantage to transduced cells, while retaining the single promoter feature of MFG responsible for high virus titer and enhanced protein production. We constructed a bicistronic retrovirus producing a single mRNA encoding both the therapeutic gene for the X-linked form of CGD (X-CGD), gp91phox, and the selectable human multidrug resistance gene, MDR1 linked together by the encephalomyocarditis virus internal ribosome entry site (IRES). As a control we constructed a bicistronic vector with the polio virus IRES element and using the bacterial neomycin resistance gene (neor) as the selective element. In Epstein-Barr virus transformed B (EBV-B) cells from an X-CGD patient, a tissue culture model of CGD, we show correction of the CGD defect and complete normalization of the cell population using either of these vectors and appropriate selection (vincristine for MDR1 and G418 for neor). Using a chemiluminescence assay of O2-. production, populations of cells transduced with either vector demonstrated initial correction levels of from less than 0.1% up to 2.7% of normal EBV-B cell oxidase activity. With either construct, cell growth under appropriate selection enriched the population of transduced cells, resulting in correction of X-CGD EBV-B cells to a level of O2-. production equalling or exceeding that of normal EBV-B cells. These studies show that a therapeutic gene can be linked to a resistance gene by an IRES element, allowing for selective enrichment of cells expressing the therapeutic gene. Furthermore, the use of MDR1 as a selective element in our studies validates an important approach to gene therapy that could allow in vivo selection and is generalizable to a number of therapeutic settings.
We studied phagocyte reduced nicotinamide adenine dinucleotide phosphate function to evaluate production of reactive oxygen species in both X-linked and autosomal forms of chronic granulomatous disease. We found a consistent and significant difference between the activated granulocyte response of the X-linked (gp91-phagocyte oxidase) form of chronic granulomatous disease (n = 18) and that of the most common autosomal recessive (p47-phagocyte oxidase) form of the disease (n = 17). The data indicate that mutations in the p47-phagocyte oxidase component of the reduced nicotinamide adenine dinucleotide phosphate oxidase component do not completely prevent oxidation despite severe defects in superoxide generation.
Early hematopoietic progenitors expressing the CD34+ phenotype can be harvested from the peripheral blood of normal individuals. We have optimized the liquid culture of human CD34+ peripheral blood progenitors (PBPs) to achieve differentiation into a population of cells consisting almost entirely of eosinophil progenitors and maturing eosinophils. Growth of CD34+ PBPs for 28 days in the presence of the combination of IL-3, granulocyte-macrophage colony-stimulating factor, and IL-5 resulted in an almost 250-fold increase in cell number, yielding a population that contained 83% maturing eosinophils. The residual population consisted of basophils and mast cells (3% by acidic toluidine blue staining, 15.2% by flow cytometric assay for binding to high-affinity IgE receptor) and immature cells. This provides an opportunity to examine the kinetics of the acquisition of specialized mature eosinophil characteristics during eosinophil differentiation. Several host-defense and bioactive proteins are found almost exclusively in eosinophil granules. In addition, stimulated eosinophils, like neutrophils, produce copious amounts of toxic oxygen radicals. We used our culture system and the sensitive technique of reverse-transcriptase polymerase chain reaction to analyze the kinetics of production of messenger RNA transcripts encoding several eosinophil proteins, including five eosinophil granule proteins and four subunit peptides of the superoxide-generating reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase in small numbers of differentiating eosinophils from peripheral blood CD34+ cells. Freshly isolated CD34+ PBPs contained transcripts for the ubiquitously present housekeeping protein phosphoglucokinase but contained no eosinophil granule protein transcripts and barely detectable amounts of some oxidase protein transcripts. On day 3 of culture, no cells recognizable by histochemical staining as eosinophils could be detected, but transcripts for all five eosinophil granule proteins were present. These transcripts increased several fold during the entire culture period. Similar kinetics were seen for all but one of the NADPH oxidase protein transcripts. However, transcripts for the p67phox NADPH oxidase protein were not detected until day 7, and functional oxidase activity did not appear until day 12. From that point, oxidase activity increased dramatically over the culture period. These studies demonstrate that commitment of CD34+ PBPs to the eosinophil lineage occurs very early, by day 3, but that further events in differentiation must take place before the appearance of histologically staining eosinophil granules and acquisition of functional oxidase capacity.
Chronic granulomatous disease (CGD) is a rare recessive disorder caused by defects in the NADPH oxidase enzyme complex of phagocytes (neutrophils, eosinophils and monocytes). CGD phagocytes fail to produce superoxide and other reactive oxygen species following cell activation (Malech, 1993). The products of oxidase activation can be measured in individual cells by flow cytometry using specific fluorescent probes that increase fluorescence upon oxidation (Trinkle et al., 1987). This approach can be used to confirm a diagnosis of CGD, and to detect the normal/abnormal phagocyte mixture that characterizes the X-linked CGD carrier state. Three fluorescent probes have been described as useful for this purpose: 2'7'-dichlorofluorescin diacetate (DCF) (Bass et al., 1983), 5,6-carboxy-2'7'-dichlorofluorescein diacetate, bis(acetoxymethyl) ester (C-DCF) (Hockenbery et al., 1993) and dihydrorhodamine 123 (DHR) (Rothe et al., 1988; Kinsey et al., 1987). A direct comparison between these three probes has not been reported. In this study we performed a direct comparison between these three probes, evaluating their ability in flow cytometric analysis to maximize fluorescent separation between activated CGD patient and normal granulocytes. Using a whole blood technique with phorbol myristate acetate (PMA) as an activator, it was found that DHR loaded normal granulocytes had a fluorescence intensity which, upon activation, was 48-fold higher than that of C-DCF loaded granulocytes and seven-fold higher than DCF loaded granulocytes (P < 0.001). Use of sodium azide to decrease the catabolism of H2O2 enhanced the fluorescence of DCF by 140%, C-DCF by 45% and DHR by 25%, suggesting that DCF is primarily sensitive to H2O2. DCF and DHR were then evaluated for sensitivity in the detection of small percentages of normal cells in a CGD/normal granulocyte mixture. Normal sub-populations as small as 0.1% could clearly be distinguished using DHR, while DCF was insensitive at this level. Based on these findings, we used DHR in an effort to detect normal granulocytes in a CGD patient following therapeutic granulocyte transfusion. We were able to detect normal granulocytes in the circulation for up to 18 h after transfusion. With these data we show that DHR is the most sensitive flow cytometric indicator for the detection of oxygen reactive species in activated granulocytes and is the best probe for evaluating CGD patients and carriers. In addition, our data suggest that DHR is a useful tool for monitoring circulating normal granulocytes in CGD patients following transfusion, and potentially will be a sensitive probe for assessing the success of such future technologies as gene therapy for CGD.
Small numbers of CD34+ primitive hematopoietic progenitors are found in normal human peripheral blood. These cells differentiate to myeloid or lymphoid lineage under the influence of different growth factors. We investigated the effects of IL5 and other growth factors on the production of eosinophils from peripheral blood CD34+ cells. CD34+ cells were enriched from normal donors by apheresis and positive selection using an affinity column and plated in agarose with different combinations of cytokines. At 14 days of growth a triple stain technique was used to identify eosinophil, monocyte, and neutrophil colonies. IL5 alone did not support colony growth from CD34+ cells. In contrast, GM-CSF and IL3 alone or together without added IL5 supported the generation of more than 50% pure eosinophil colonies. Addition of IL5 did not change the total number of colonies, but increased the fraction of pure eosinophil colonies to over 70%. Addition of G-CSF reduced the percentage of eosinophil colonies and increased the percentage of neutrophil colonies. Under the best conditions for eosinophil colony growth (IL3+GM-CSF+IL5), the addition of interferon-alpha or bacterial lipopolysaccharide inhibited colony growth by 51 and 58%, respectively. Addition of interferon-gamma, tumor necrosis factor-alpha, or dexamethasone had no effect on eosinophil colonies. Since IL5 alone did not support colony growth from CD34+ cells, we determined when IL5-responsive cells appeared in culture. Cells were grown initially with IL3 + GM-CSF in suspension, washed, and plated in agarose with IL5 alone. Only when progenitors were grown at least 3 days could IL5 serve as the single growth factor supporting pure eosinophil colony growth (47 colonies/10(4) cells plated at Day 3 and 134 colonies/10(4) cells at Day 7). We used neutralizing anti-IL5 antibodies to demonstrate that this late acting IL5 growth effect was specific, and that differentiation of eosinophils in the presence of IL3 + GM-CSF was IL5 independent. Using reverse-transcriptase polymerase chain reaction, the mRNA encoding the eosinophil-specific protein eosinophil peroxidase (EPO) was not detected in Day 0 CD34+ cells, but was demonstrated by Day 3 of culture. We conclude that within 3 days of culture, peripheral blood CD34+ cells can become committed to the eosinophil lineage as demonstrated by responsiveness to IL5 and production of EPO transcripts.
Recombinant human granulocyte colony-stimulating factor (G-CSF) treatment has been shown to increase average neutrophil counts substantially in patients with childhood-onset cyclic neutropenia (or "cyclic hematopoiesis"), but not to eliminate the cyclic oscillations of neutrophil counts or those of other blood elements (monocytes, platelets, eosinophils, and reticulocytes) that are characteristic of this hematopoietic disorder. Indeed, oscillations of neutrophil counts are amplified during G-CSF treatment. We have compared the effects of recombinant granulocyte-macrophage-CSF (GM-CSF) with those of G-CSF in three patients with this disease (2 men and 1 woman, 17, 30, and 32 years of age). These patients were treated with GM-CSF (2.1 micrograms/kg/day, subcutaneously) for 6 weeks, preceded and followed by 6 to 13 weeks of detailed observation to document changes in the cyclic oscillations of blood neutrophils and other blood elements; two of the patients were subsequently treated with G-CSF (5.0 micrograms/kg/d, subcutaneously) and observed for comparable periods of time. Unlike G-CSF treatment, which increased average neutrophil counts more than 20-fold, GM-CSF increased neutrophil counts only modestly, from 1.6- to 3.9-fold, although eosinophilia of varying prominence was induced in each patient. However, at the same time, GM-CSF treatment dampened or eliminated the multilineage oscillations of circulating blood elements (neutrophils, monocytes, platelets, and/or reticulocytes) in each of the patients. In contrast, G-CSF treatment of the same patients markedly amplified the oscillations of neutrophil counts and caused the cycling of other blood elements (monocytes in particular) to become more distinct. These findings support the conclusion that the distinctive cycling of blood cell production in childhood-onset cyclic neutropenia results from abnormalities in the coordinate regulation of both GM-CSF-responsive, multipotential progenitor cells and G-CSF-responsive, lineage-restricted, neutrophil progenitors.
Chronic granulomatous disease (CGD) can result from any of four single gene defects involving components of the superoxide (O2-.)-generating phagocyte NADPH oxidase (phox). The phox transmembrane flavocytochrome b558 is composed of two peptides, gp91phox and p22phox. Mutations of gp91phox cause X-linked CGD, whereas mutations of p22phox cause one of the three autosomal recessive forms of CGD. We used the Maloney leukemia virus-based MFG retrovirus vector to produce replication defective retroviruses encoding gp91phox or p22phox. To maximize viral titer MFG retroviruses do not contain internal promoter or resistance elements. Epstein-Barr virus transformed B-lymphocyte cell lines (EBV-B) derived from normal individuals contain phox components and produce O2-., whereas those derived from CGD patients show the CGD defect. Transduction of gp91phox or p22phox-deficient CGD EBV-B lines resulted in correction of O2-. production from a barely detectable baseline to an average 7.2% and 13.8% of normal control, respectively, without any selective regimen to enrich for transduced cells. CD34+ hematopoietic progenitor cells, the therapeutic target for gene therapy of CGD, were isolated from peripheral blood of CGD patients, transduced with MFG-phox retroviruses, and differentiated in culture to mature phagocytes. Transduction of progenitors corrected the gp91phox (seven patients) and p22phox (two patients) CGD phagocyte oxidase defect to 2.5% and 4.9% of normal O2-. production, respectively, representing an 87-fold and 161-fold increase. These studies show correction of flavocytochrome b558-deficient CGD in primary hematopoietic progenitors, providing a basis for development of gene therapy for the X-linked gp91phox and autosomal p22phox-deficient forms of CGD.
The phagocyte superoxide-generating NADPH oxidase, a multicomponent, membrane-bound electron transport chain, consists of cytochrome b558, p47-phox, p67-phox, and p21rac1 or p21rac2. The mechanisms of oxidase assembly are poorly understood. In previous studies using a cell-free NADPH oxidase system, we showed that preincubation of neutrophil membrane with neutrophil cytosol containing p47-phox, but not p67-phox, led to formation of a long-lived NADPH oxidase intermediate. This suggested that p47-phox interacted with cytochrome b558 in the early stages of oxidase assembly while p67-phox participated in a later stage. Peptides containing the sequence RGVHFIF (corresponding to amino acids 559-565 of the 91-kDa subunit of cytochrome b558) inhibit NADPH oxidase activity by blocking the early interaction between p47-phox and cytochrome b558. In the present study, we examined whether p21rac facilitated the interaction between p47-phox and cytochrome b558. We preincubated pure recombinant p47-phox with neutrophil membrane containing cytochrome b558 in the cell-free system. Superoxide-generating activity was subsequently reconstituted by adding pure rp67-phox and partially purified p21rac. RGVHFIF inhibited superoxide production if added to the cell-free system during preincubation of rp47-phox with membrane. RGVHFIF was markedly less inhibitory if added to the cell-free system after membrane was preincubated with pure rp47-phox. In contrast to p47-phox, preincubation of membrane with either p21rac or rp67-phox conferred no protection from inhibition of superoxide-generating activity by RGVHFIF added after preincubation.(ABSTRACT TRUNCATED AT 250 WORDS)